SYSTEM AND METHOD FOR IDENTIFYING NUCLEAR THREATS

Abstract

A method and a device for the detection of radioactive sources, based on the simultaneous use of two or more radiation detectors of different types and the composition of the data collected by the two or more radiation detectors.

Claims

1. A method for identifying a radioactive source comprising the steps of: detecting, from said radioactive source, gamma emitting sources with a gamma spectroscopic, type I, detector; detecting, from said radioactive source, the plurality of impulses associated with respective neutron emissions and gamma emissions with a gamma/neutron, type II, detector; identifying a first portion of said plurality of impulses associated with respective neutron emissions and a second portion of said plurality of impulses associated with respective gamma emissions, said step of identifying said first and second portion of impulses including the calculation for each impulse of a respective total integral Ltot of the impulse, and further comprising the steps of: determining a first number n.sub.n of impulses belonging to the first portion and a second number n.sub.γ of impulses belonging to the second portion; calculating a first average value E.sub.n and a second average value E.sub.γ of the total integral Ltot, wherein E.sub.n=Σ.sub.Nn (Ltot/n.sub.n), and E.sub.γ=Σ.sub.Ng (Ltot/n.sub.γ); calculating parx=n.sub.γ/n.sub.n and pary=E.sub.γ/E.sub.n; identifying if the radioactive source belongs to a first group of neutron sources or to a second group of gammas sources by the data parx, pary; if the radioactive source belongs to the first group of neutron sources, perform identification of the gamma sources detected with a type I detector, through spectroscopy; double confirmation of the identification if both gamma and neutron are detected; detection of masking condition of neutrons in presence of gamma, performed by merging the results of the identification of the Type I detector with the results of the identification of the Type II detector; detection and identification of gamma and neutron sources in presence of moderators and/or shielding materials through the Type II detector.

2. The method according to claim 1, wherein detecting, from said radioactive source, of gamma emitting sources with a gamma spectroscopic, type I, detector comprises acquiring: a first identification spectrum, Spec1; a background as a reference spectrum; a second identification spectrum, Spec2=Spec1—Background.

3. The method according to claim 1, wherein detecting, from said radioactive source, of gamma emitting sources with a gamma/neutron, type II, detector comprises acquiring: a third identification spectrum, Spec3; a fourth identification spectrum, Spec4; a background as a reference spectrum.

4. The method according to claim 2, wherein detecting, from said radioactive source, of gamma emitting sources with a gamma spectroscopic, type I, detector comprises: searching for gamma peaks on the second identification spectrum, spec2; validation of spec2 peaks; validation of spec1 peaks; searching for further peaks; computation of integrals within a Region Of Interest; searching of multiplets; rejection of possible duplicated peaks; Compton edge and backscatter peak identification; gamma source identification; minor source identification;

5. The method according to claim 1 wherein merging the results of the identification of the Type I detector with the results of the identification of the Type II detector comprises: validation of AmLi source; masking condition determination for neutron sources; estimation of enrichment grade.

6. The method according to claim 1, wherein the type II detector is replaced by a thermal neutron, type III detector.

7. The method according to claim 6, wherein the type I detector is replaced by a type II, detector.

8. The method according to claim 1, wherein the type I detector uses Cerium-Bromide (CeBr.sub.3) crystal lattice.

9. The method according to claim 1, wherein the type II detector uses xylene as an organic solvent.

10. A device for identifying a radioactive source comprising: a type II detector (3); a first photomultiplier or SiPm readout (4) associated to said type II detector (3); a first digitizer circuit (5) associated to said first photomultiplier (4) and adapted to convert light pulses into digital signals; a first power supply circuit (6) associated to said first photomultiplier (4); a type I detector (3-bis); a second photomultiplier or SiPm readout (4-bis) associated to said type I detector (3-bis); a second digitizer circuit (5-bis) associated to said second photomultiplier (4 -bis) and adapted to convert light pulses into digital signals; a second power supply circuit (6-bis) associated to said second photomultiplier (4-bis); a programmable digital processor (9) associated to said first and second photomultiplier (4, 4-bis); a user interface, GUI, (7); and a control panel (8) for the exchange of information and data with a user.

11. The device according to claim 10, wherein the programmable integrated circuit is programmed with a plurality of instructions for execution of a method for identifying a radioactive source comprising the steps of: detecting, from said radioactive source, gamma emitting sources with a gamma spectroscopic, type I, detector; detecting, from said radioactive source, the plurality of impulses associated with respective neutron emissions and gamma emissions with a gamma/neutron, type II, detector; identifying a first portion of said plurality of impulses associated with respective neutron emissions and a second portion of said plurality of impulses associated with respective gamma emissions, said step of identifying said first and second portion of impulses including the calculation for each impulse of a respective total integral Ltot of the impulse, and further comprising the steps of: determining a first number n.sub.n of impulses belonging to the first portion and a second number n.sub.γ of impulses belonging to the second portion; calculating a first average value E.sub.n and a second average value E.sub.γ of the total integral Ltot, wherein E.sub.n=Σ.sub.Nn (Ltot/n.sub.n), and E.sub.γ=Σ.sub.Ng (Ltot/n.sub.γ); calculating parx=n.sub.γ/n.sub.n and pary=E.sub.γ/E.sub.n; identifying if the radioactive source belongs to a first group of neutron sources or to a second group of gammas sources by the data parx, pary; if the radioactive source belongs to the first group of neutron sources, perform identification of the gamma sources detected with a type I detector, through spectroscopy; double confirmation of the identification if both gamma and neutron are detected; detection of masking condition of neutrons in presence of gamma, performed by merging the results of the identification of the Type I detector with the results of the identification of the Type II detector; detection and identification of gamma and neutron sources in presence of moderators and/or shielding materials through the Type II detector.

12. The device according to claim 10, wherein the type I detector uses Cerium-Bromide (CeBr.sub.3) crystal lattice.

13. The device according to claim 10, wherein the type II detector uses xylene as an organic solvent.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0028] Further features and advantages of the present invention will be apparent from reading the following detailed description, provided by way of non-limiting example, with the aid of the figures illustrated in the attached tables, in which:

[0029] FIG. 1A illustrates a portion of the flowchart of the method for gamma and neutron identification according to the invention that relates to the searching for gamma peaks;

[0030] FIG. 1B illustrates the step of validation of peaks from FIG. 1A;

[0031] FIG. 1C illustrates an intermediate step of validation of additional peaks from FIG. 1A;

[0032] FIG. 2A illustrates a further intermediate step of validation of other peaks from FIG. 1C;

[0033] FIG. 2B illustrates intermediate steps of identifying the integrals within the Region of Interest (ROI) and multiplets searching:

[0034] FIG. 2C illustrates the steps for the rejection of possible duplicate peaks, and Compton edge and backscatter peak identification following steps illustrated in FIG. 2B;

[0035] FIG. 3A illustrates the step of source identification following identification steps illustrated in FIG. 2C;

[0036] FIG. 3B illustrates the steps of minor source identification and the validation of Am/Li source following the steps illustrated in FIG. 3A;

[0037] FIG. 3C illustrates the steps for masking condition determination for neutron sources following the steps illustrated in FIG. 3B;

[0038] FIG. 4A illustrates the steps for identifying specified elements;

[0039] FIG. 4B illustrates further steps for identifying elements following the steps of FIG. 4A; and

[0040] FIG. 4C illustrates the steps for the estimation of enrichment grade and concludes the identification.

[0041] FIG. 5 illustrates a chart of SNM identification experimental results with no masking and no shielding wherein each point represents a 1 min measurement. The clusters of the measurements fall in the center of the rectangular regions each corresponding to a specific SNM;

[0042] FIG. 6 illustrates a chart of Cf-252 identification experimental results with lead shield and no masking;

[0043] FIG. 7 illustrates a chart of Cf-252 identification experimental results with polyethylene (PE) shield and no masking;

[0044] FIG. 8 illustrates a chart of Cf-252 identification experimental results with masking scenarios carried out with gamma sources with different intensities;

[0045] FIG. 9 illustrates a chart of the total results of the identification of all the SNMs in different measurement conditions. Points in blue, orange, green, red and purple respectively represent the measurements with sources of Cf-252, Pu, U, AmBe and AmLi;

[0046] FIG. 10 illustrates the same chart of FIG. 9 with more than 8 neutrons per minute;

[0047] FIG. 11 illustrates the same chart of FIG. 9 with more than 8 neutrons per minute without shielding;

[0048] FIG. 12 illustrates a chart of repeated measurements n-times with two sources of PU with different enrichments at 60% and 70%;

[0049] FIG. 13 illustrates a chart showing Pu with different degrees of enrichment without shielding and masking;

[0050] FIG. 14 illustrates a chart of a selection from the chart of FIG. 9 showing identifications with sources shielded by Pb;

[0051] FIG. 15 illustrates a chart of a selection from the chart of FIG. 9 showing identifications with sources shielded by PE;

[0052] FIG. 16 illustrates a chart of a selection from the chart of FIG. 9 showing identifications with sources shielded by Pb+PE, and

[0053] FIG. 17 illustrates a chart of SNM identification in masking condition with gamma sources.

[0054] FIG. 18 illustrates a block diagram of the device for gamma and neutron identification according to the invention.

[0055] The following description of exemplary embodiments refers to the accompanying drawings. The same reference numerals in different drawings identify the same or similar elements. The following detailed description does not limit the invention. The scope of the invention is defined by the attached claims.

[0056] In the following description reference numbers 1.a) to 16.a); 1.b) to 5.b) and 1.c) to 8.c) which are used in the flowcharts of FIG. 1A to FIG. 4C have been inserted in the text between square brackets.

DETAILED DESCRIPTION OF THE INVENTION

[0057] The present invention concerns a method and an device for the detection of radioactive sources, based on the simultaneous use of two or more radiation detectors of different types and in particular consisting of a possible composition of detectors of type I, II or III In further detail, the detectors used simultaneously according to one embodiment of the present invention are:

[0058] Type I detectors capable of producing a gamma spectrum which can be used in the system of the present invention. These detectors are organic or inorganic scintillators based on luminescent materials and coupled with a signal readout system (for example but not limited to PMT or SiPM) type devices. The luminescent material can be chosen between NaI(Tl), NaIL, CsI, LaBr.sub.3, CLLB, BaF.sub.2, CdWO.sub.4, CaF.sub.2(Eu), CaWO.sub.4, CdWO.sub.4, Gd.sub.2O.sub.2S, LaCl 3(Ce), PbWO.sub.4, LuI.sub.3, LSO, LYSO, YAG(Ce), ZnS(Ag), ZnWO.sub.4, BGO, and CeBr.sub.3 etc. Type I detectors could be also high-resolution semiconductor detectors such as those with hyper-pure Germanium and, in particular, HpGe, CdZnTe, TlBr, etc.

[0059] Type II detectors capable of counting both gamma and neutrons with adequate and properly designed processing electronics (typically digitizers but not limited to) of their signals, such as liquid scintillator or plastic scintillator detectors, both organic or inorganic. An example is the organic liquid scintillator detector EJ309 used by us for validation but not limited to. In fact, other similar scintillation detectors could be used with scintillation proprieties enough to allow to resolve gamma and neutron pulses from the same detector and correlate these pulses with a very precise timing up to 10 ns if combined with a properly designed digitizers using PSD Pulse Shape Discrimination.

[0060] Type III Detectors with thermal neutron detection capabilities or gas detectors, including gas, solid state or plastic scintillator detector with high efficiency for the detection of thermal neutrons.

[0061] The combination of two detectors, for instance arbitrarily chosen among one of type I and one of type II allows:

[0062] For type I: Obtaining a gamma spectrum in an energy range from 0 KeV to 10 MeV.

[0063] For type II: Obtaining a gamma spectrum with low resolution but enough to allow seeing the Compton edge of the radiation to which they are subjected and simultaneously to detect neutrons with enough efficiency.

[0064] The method according to the present invention is based on calculation algorithms that allow detecting radioactive sources by exactly identifying their type and not merely their generality.

[0065] One preferred embodiment of the method for identifying a radioactive source according to the present invention makes use of the following data collected from said type I and type II detectors:

[0066] Data from Type I detector (gamma spectroscopic detectors) that generates:

[0067] Spec1: identification spectrum acquired with Type I detector in case of gamma alarm.

[0068] Background-Type I—acquired with Type I detector continuously when no alarm is present to have a reference spectrum in absence of alarm.

[0069] Spec2=Spec1—Background-Type I normalized by acquisition duration.

[0070] Data from Type II detector (gamma/neutron detectors) that generates:

[0071] Spec3: gamma identification spectrum acquired with Type II detector in case of only neutron or gamma and neutron alarm.

[0072] Spec4: neutron identification spectrum acquired with Type II detector in case of only neutron or gamma and neutron alarm.

[0073] Background-Type II—acquired with Type II detector continuously when no alarm is present to have a reference spectrum.

[0074] Furthermore, one preferred embodiment of the method for identifying a radioactive source according to the present invention comprises the steps of:

[0075] Detecting gamma emitting sources with a type I detector.

[0076] Detecting the plurality of impulses originating from a type II detector associated with respective neutron emissions and gamma emissions from said radioactive source.

[0077] identifying a first portion of said plurality of impulses associated with respective neutron emissions and a second portion of said plurality of impulses associated with respective gamma emissions, said step of identifying said first and second portion of impulses including the calculation for each impulse of a respective total integral Ltot of the impulse;

[0078] determining a first number n.sub.n of impulses belonging to the first portion (total number of events in Spec4) and a second number n.sub.γ of impulses belonging to the second portion (total number of events in Spec3);

[0079] calculating a first average value E.sub.n and a second average value E.sub.γ of the total integral L.sub.tot, in which

[00001]$E_n={.Math.}_{N_n}\frac{L_{tot}}{N_n}\mathrm{and}$$E_{\gamma }={.Math.}_{N_g}\frac{L_{tot}}{N_g};$

[0080] Then calculating a datum par.sub.x=n.sub.γ/n.sub.n and a datum par.sub.γ=Ē.sub.γ/Ē.sub.n;

[0081] identifying as a function of the data par.sub.x, par.sub.γ if the radioactive source belongs to a first group of neutron sources or to a second group of gammas sources, wherein if the radioactive source belongs to the first group of neutron sources, identifying that radioactive source through the following steps:

[0082] perform identification of the gamma sources detected with a type I detector, through spectroscopy.

[0083] merge data coming from type I and type II detectors through the following steps:

[0084] step 1: double confirmation of the Special Nuclear Material SNM identification if both gamma and neutron are detected. In an embodiment, the Type I detector can identify the SNM sources through a gamma spectroscopy measurement consisting in the identification of the characteristic gamma lines emitted by the sources. Moreover, an enrichment level estimation of the SNM is performed by measuring the ratio of the detected gamma lines and their respective backgrounds. The second confirmation is provided by the Type II detector identification that weight the gamma and neutron counting and energy ratios (parx and pary) to identify the SNM source.

[0085] Step 2: detection of masking condition of neutrons in presence of gamma, performed by merging the results of the identification of the Type I detector with the results of the identification of the Type II detector. In fact, the Type I detector is sensitive only to gamma radiation while Type II detector can detect both gamma and neutron. At the end of the identification procedure the results are merged and if additional gamma sources are detected in addition to the detected SNMs, then the masking condition is confirmed.

[0086] Step 3: detection and identification of gamma and neutron sources in presence of moderators and/or shielding materials through the Type II detector.

[0087] The identification algorithm of the Type II detector provides identification regions for each SNM defined by values of par.sub.x and par.sub.y. For each region of each SNM it is possible to define sub-areas obtained from the variation of par.sub.x, par.sub.y or both according to the variation of the gamma or neutron measured rate. The gamma and neutron rate variation are strictly correlated to the presence of heavy material shield or moderator material that cause the shift of the par.sub.x and par.sub.y value.

Gamma Spectrum Correction from Contribution of Masking Gamma Sources

[0088] The use of an inorganic scintillator detector, such as CeBr.sub.3, of type I, allows the result of the identification method of neutron emitters to be corrected in the event of a high range end or the presence of masking sources.

[0089] The gamma spectrum can in some conditions include both radiation from the neutron emitting source and from any masking sources. The masking source can therefore modify the result of the identification of the neutron source.

[0090] The use of type I detector allows gamma sources to be identified through their energy spectrum and gamma line emission. Furthermore, starting from the energy of the gamma radiation of the identified masking source and from the knowledge of the Klein-Nishina distribution for each gamma emission, it is possible to obtain normalized templates of the spectral distribution of the type II detector: KN.sub.i.sup.Isot,j where i is the spectral channel index, Isot identifies the identified isotope, j identifies the gamma emission lines of Isot.

[0091] From the count of each photopeak, from the photopeak efficiency measurement (cps/phi) of the type I detector and from a measurement of the counting efficiency of the type II detector it is possible to obtain a multiplicative factor that estimates the contamination in the gamma spectrum of the type II due to any masking sources:

[00002]$n_{Tot}^{\mathrm{EJ}309}=\frac{n_{\mathrm{Peak}}^{\mathrm{Inor}.}\times {\mathrm{eff}}_{Tot}^{\mathrm{EJ}309}}{{\mathrm{eff}}_{\mathrm{Peak}}^{\mathrm{Inor}.}}$

[0092] The coefficients eff.sub.Tot.sup.Ej309 and eff.sub.Peak.sup.Inor can be determined, for example, using the Montecarlo method and then verified or weighed through experimental measurements with some certified sources.

[0093] The sum of the distributions obtained by multiplying the term n.sub.Tot.sup.EJ309 by the corresponding normalized Klein-Nishina distribution, gives the expected gamma distribution of the masking source measured by the type II detector:

Bkg.sub.i=Σ.sub.IsotΣ.sub.jn.sub.Tot Isot,j.sup.EJ309×KN.sub.i.sup.Isot,j

[0094] Once the contamination has been estimated, it is possible to subtract its contribution in the gamma spectrum of the type II measured and perform the identification as already implemented.

[0095] The Differences between the state-of-the-art detection methods and detectors and the present invention method are summarized in the charts below:

TABLE-US-00001 Use of multi-libraries Areas/spaces Masking for gamma identification Area/space definition for correction Gamma optimization and definition by SNM in the through the use spectrum simultaneous neutron identification presence of of the second Gamma Neutron Neutron with nuclides identification confidence both shielding spectrometric count count identification identification optimization level for SNM and moderators detector State of YES YES YES but NO NO NO NO NO the art with limits in specific conditions Present YES YES YES YES YES YES YES YES invention

TABLE-US-00002 State of the art Present invention Gamma count YES YES Neutron count YES YES Neutron identification NO but with limits in YES specific conditions Gamma spectrum with nuclides identification YES YES Use of multi-libraries for gamma identification optimization and NO YES simultaneous neutron identification optimization Area/space definition by identification confidence level for SNM NO YES Areas/spaces definition for SNM in the presence of both shielding NO YES and moderators Masking correction using the second spectrometric detector NO YES

TABLE-US-00003 Gamma Gamma- detection neutron device device Present Invention Feature Gamma counting ✓ ✓ ✓ 1 s gamma alarm for minimum dose rate of 500 nSv/h 500 nSv/h 50 nSv/h Gamma identification in 1 minute ✓ ✓ ✓ Standard resolution of gamma detector >6% >6% <3.5% Max number of sources identified simultaneously From 3 to 7 From 3 to 7 Gamma up to 7 gamma gamma neutron up to 2 sources sources Max number of peaks detected correctly in gamma 3-12 3-12 Up to 16 ID Neutron counting X ✓ ✓ 1 s neutron alarm for 20.000 n/s of Cf-252 at a X 25 cm 125 cm distance of 1 s neutron alarm for 20.000 n/s of Pu-239 at a X 25 cm 500 cm distance of 1 s neutron alarm for 20.000 n/s of Cf-252 with a X 5 cm 10 cm + 5 cm shielding of HDPE of lead 1 s neutron alarm for 20.000 n/s of Pu-239 with a X 5 cm 10 cm + 5 cm shielding of HDPE of lead Identification of neutron source in 1 minute X X ✓ Identification of SNM with shielding, masking, or X X ✓ moderation Application example Detection of common gamma sources (Co-60, Cs- ✓ ✓ ✓ 137, I-131) Identification of common gamma sources (Co-60, ✓ ✓ ✓ Cs-137, I-131) Detection of SNM (gamma and neutron emitting ✓ gamma ✓ gamma ✓ gamma material like U, Pu, Am/Be . . .) X neutron ✓ neutron ✓ neutron ID of SNM (gamma and neutron emitting material ✓ gamma ✓ gamma ✓ gamma like U, Pu, Am/Be . . .) X neutron X neutron ✓ neutron Detection of SNM with only n or with shielded X gamma X gamma X gamma gamma (like Cf-252, U, Pu, Am/Be in a lead box) X neutron ✓ neutron ✓ neutron extra info: shielded ID of SNM with only n or with shielded gamma (like X gamma X gamma X gamma Cf-252, U, Pu, Am/Be in a lead box) X neutron X neutron ✓ neutron ID of SNM like previous point with masking with ✓ I-131 ✓ I-131 ✓ I-131 common gamma source (like I-131) X SNM X SNM ✓ SNM extra info: masked

[0096] The flow diagrams of enclosed FIG. 1A-FIG. 4C show a flow chart of an embodiment of the method according to the present invention starting from the alarm event, the detection of a neutron source.

[0097] The method according to the present description is adapted to work correctly with any type I detector. Testing has been performed with many detectors, the measurement results and parametrization reported herein are related to the average resolution scintillator CeBr3.

[0098] In an embodiment, the method according to the present description makes use of 2 gamma energy spectra [1.a)]: A first spectrum, spec.sub.1 is obtained during the identification period (for example during 1 minute of data acquisition, but other durations can be chosen) and a second spectrum, spec.sub.2 is equal to spec.sub.1, with the subtraction of last environmental background (for example after 3 minutes of data acquisition but other durations can be chosen), properly rescaled. In addition, a Background is computed, and it is subtracted to spec.sub.2 [2.a)]. An example of the algorithm steps is described hereinafter, wherein the values of the parameters and formulas employed are related to the specific choice of a CeBr3 type I detector. In general, these formulas and parameters can be changed depending on the choice of the type I detector. For instance, peak fit can be performed with Gaussian or Poisson fit, background can be calculated with a polynomial formula form degree zero to degree three or with a step-function.

[0099] 1) searching for gamma peaks on spec.sub.2

[0100] This can be done using a Search function [3.a)].This function restricts the number of possible peaks through two parameters, σ and thr. The first one selects peaks based on their width (the smaller σ, the smaller peak width will be accepted). The second one discards peaks with amplitude less than thr.Math.h, with h equal to the height of the highest peak. σ is set to 2, 5, and 7 spectrum channels (1 channel=3 keV) respectively for the energy intervals [0, 330[, [330, 1100[and [1100, 5100[keV. thr is set to 0.0002.

[0101] 2) 1.sup.st validation of spec.sub.2 peaks.

[0102] To be performed for each peak position, p.sub.1,i, found at step 1). Fitting, for example through Gaussian fit [4.a)] is performed within the energy interval [a.sub.i; b.sub.i], with a.sub.i=p.sub.1,i−5/3 σ.sub.t(p.sub.1,i), b.sub.i=p.sub.1,i+5/2σ.sub.t(p.sub.1,i) and σ.sub.t(p.sub.1,i) the theoretical resolution value, expressed in standard deviation, at the energy p.sub.1,i (if (b.sub.i−a.sub.i)<40 keV, then a.sub.i=p.sub.1,i−11 keV and b.sub.i=p.sub.1,i+10 keV). Standard deviation, σ.sub.i, centroid, E.sub.i and Gaussian integral, Int.sub.i, are retrieved from the fit, discarding immediately those peaks with non-physical values (E.sub.i or Int.sub.i negative, E.sub.i<a.sub.i, E.sub.i>b.sub.i, E.sub.i>5100 keV) or with Int.sub.i<30 counts [5.a)].

[0103] Then peak energy resolution, FWHM.sub.i=2√{square root over (2 ln(2))}.Math.σ.sub.i, is checked. Peak at energy E.sub.i is discarded in the following cases:

FWHM.sub.i>FWHM.sub.max(E.sub.i) or FWHM.sub.i<FWHM.sub.min(E.sub.i)

with

FWHM.sub.max(E.sub.i)=1.65.Math.FWHM.sub.t(E.sub.i)

FWHM.sub.min(E.sub.i)=0.56.Math.FWHM.sub.t(E.sub.i)

[0104] An offset equal to 0.03.Math.E.sub.i is added to FWHM.sub.max(E.sub.i) if E.sub.i<70 keV. FWHM.sub.t(E) is the theoretical resolution value (expressed as FWHM [keV]) at energy E, equal to

FWHM.sub.t(E)=√{square root over (−4.563094418.Math.10.sup.−5E.sup.2+1.203082155E+9.365215638)}

[0105] Clearly σ.sub.t(E) is equal to FWHM.sub.t (E)/2√{square root over (2 ln(2))}.

[0106] 3) 2.sup.nd validation of spec.sub.1, peaks. Gamma spectrum is switched from spec.sub.2 to spec.sub.1 and a second validation is performed on the gamma energy spectra spec.sub.1. A fitting procedure is performed [6.a)], for instance with a function defined as

f(E)=Gauss(E;E.sub.i,f,σ.sub.i,f,Int.sub.i,f)+m.Math.E+q

[0107] within the energy interval [E.sub.1,i, E.sub.2,i,], with E.sub.1,i=E.sub.i−3.5 σ.sub.t(E.sub.i) and E.sub.2=E.sub.i+3.5 σ.sub.t(E.sub.i). In the fitted function, f(E), the Gaussian stands for the photopeak while the last term, m.Math.E+q, is a linear function for the estimation of the local background. Fitting parameters m, q, σ.sub.i,f, E.sub.i,f are initialized in the following way: [0108] m=(c.sub.2,i−c.sub.1,i)/(E.sub.2,i−E.sub.1,i), with c.sub.2,i and c.sub.1,i as the spectrum counts at the energies E.sub.2,i and E.sub.1,i, respectively; [0109] q=c.sub.1,i−(m.Math.E.sub.1,i); [0110] σ.sub.i,f, E.sub.i,f with the parameters got from the previous fit, σ.sub.i and E.sub.i.

[0111] Furthermore, assuming y.sub.1=m.Math.E.sub.1+q and y.sub.2=m.Math.E.sub.2+q, if y.sub.1 or y.sub.2 are negative the fitting procedure is iterated for a maximum of 5 attempts, forcing fitting parameters to these constraints:

[0112] 1. if y.sub.1<0 and y.sub.2<0.fwdarw.m=0, q=0;

[0113] 2. if y.sub.1<0 and y.sub.2≥0.fwdarw.max(m)=1.01.Math.y.sub.2/(E.sub.2−E.sub.1) counts/keV, both m and q initialized to 0;

[0114] 3. if y.sub.1≥0 and y.sub.2<0.fwdarw.min(m)=−1.01.Math.y.sub.1/(E.sub.2−E.sub.1) counts/keV, both m and q initialized to 0; Peaks with a negative value of Int.sub.i,f or with E.sub.i,f outside [E.sub.1,i, E.sub.2,i] are immediately discarded. At this point [7.a)], the integral, S.sub.i, of the Gaussian Gauss (E; E.sub.i,f, σ.sub.i,f, Int.sub.i,f) and the integral, B.sub.i, of the linear background m.Math.E+q are computed within the energy interval [E.sub.min,i, E.sub.max,i], with E.sub.min,i=E.sub.i,f−3.5 σ.sub.i,f and E.sub.max,i=E.sub.i,f+3.5 σ.sub.i,f. These two values are used to reject false peaks, that would otherwise be detected due to statistical fluctuations. Another parameter called “significance”, Sig.sub.i, is computed. This parameter is equal to S.sub.i/√(S.sub.i+B.sub.i), that is the ratio between the signal integral, S.sub.i, and the statistical uncertainty of total counts within [E.sub.min,i, E.sub.max,i]: peaks with Sig.sub.i<6 are rejected [8.a)]. It is also computed the parameter Sig/ch (it will be used later, see point 5), that is the ratio between Sig and the number of spectrum channels included in same energy range [E.sub.min,i, E.sub.max,i].

[0115] Finally the last check on peak energy resolution, FWHM.sub.i,f=2√{square root over (2 ln(2))}σ.sub.i,f/E.sub.i,f, rejects those peaks with FWHM.sub.i,f>FWHM.sub.max(E.sub.i,f) or FWHM.sub.i,f<FWHM.sub.min(E.sub.i,f), where

FWHM.sub.max(E.sub.i,f)=1.25.Math.FWHM.sub.t(E.sub.i,f)

FWHM.sub.min(E.sub.i,f)=0.60.Math.FWHM.sub.t(E.sub.i,f)

[0116] An offset value equal to 0.07.Math.E.sub.i,f is added to FWHM.sub.max(E.sub.i,f) if E.sub.i,f<70 keV.

[0117] 4) Searching for further peaks (second cycle of steps 2 and 3). First of all, every Gaussian function (that one at energy E.sub.i, got at point 2) related to all gamma peaks previously not rejected is subtracted 1 from both spec.sub.1 and spec.sub.2 (subtraction within the energy interval [0.5.Math.E.sub.min,i, . . . , 1.5.Math.E.sub.max,i]) [9.a)].Then points 1), 2), and 3) are repeated using the updated spectra, spec.sub.1 and spec.sub.2, and a different set of parameters for the Search function. Now σ is set to 10 and 12 spectrum channels, respectively for the energy intervals [0, 1100[and 11100, 5100[keV, and thr is set to 0.006.

[0118] The reason for this operation is to improve the identification capability of the algorithm. The subtraction of the peaks makes it possible to also recognized some peaks that would not be separated enough (e.g., .sup.22Na 1275 keV and .sup.60Co 1332 keV, see FIG. 1).

[0119] 5) Integrals within the Region of Interest (ROI) 180-243 keV. Two integrals, P.sub.208 and B.sub.208, are computed within the energy interval [180 keV, 243 keV] to possibly evaluate .sup.239Pu enrichment grade [4.c)] (higher values of P.sub.208/B.sub.208 should indicate the presence of a greater contribution from .sup.241Pu isotope). P.sub.209 and B.sub.208 are computed as the number of total counts within the ROI using spec.sub.2 and Background (the same one subtracted to spec.sub.2, at the beginning), respectively [10.a)].

[0120] 6) Multiplets searching. For each peak at energy E.sub.i,f detected at the end of 4), the algorithm looks for those peaks, at energy E.sub.j,f(with E.sub.i,f>E.sub.j,f), such that E.sub.i,f−E.sub.j,f<6.5.Math.σ.sub.i,f. In this case [11.a)], both 2 peaks at E.sub.i,f and E.sub.j,f are discarded and point 3) is repeated analogously. The only difference is the fitting function, defined as

f(E)=Gauss.sub.i(E;E.sub.i,f,σ.sub.i,f,Int.sub.i,f)+Gauss.sub.j(E;E.sub.j,f,σ.sub.j,f,Int.sub.j,f)+m.Math.E+q

[0121] Function parameters of both Gauss.sub.i and Gauss.sub.j are initialized with the respective ones got at the end of the fitting procedure described at point 2), while m and q are initialized as described in 3), with the fitting energy interval equal to [E.sub.1,i, E.sub.2,j]. Also the validation procedure is the same one described at 3) for both peaks at E.sub.i,f and E.sub.j,f.

[0122] 7) Rejection of possible duplicated peaks. Each couple of peaks whose energies, E.sub.i,f and E.sub.j,f respectively (with E.sub.i,f>E.sub.j,f), differ by less than 1.0.Math.σ.sub.i,f, is discarded. These two peaks are replaced by a new peak whose parameters (E.sub.i,f, σ.sub.i,f, Int.sub.i,f, Sig, Sig/ch) are computed as the average of the respective parameters of the two peaks discarded [12.a)];

[0123] 8) Compton edge and backscatter peak identification. Theoretical Compton edge, E.sub.C, and backscatter peak, E.sub.BS, are computed for each identified peak with energy E.sub.i,f>200 keV. Then, for each peak at the energy E.sub.j,f (E.sub.j,f<E.sub.i,f), assuming σ.sub.t(E.sub.j,f) as its theoretical resolution value (σ.sub.t(E)=FWHM.sub.t(E)/2.354820045), its distances from E.sub.C and E.sub.BS are respectively evaluated as

d.sub.C=|(E.sub.j,f+σ.sub.t(E.sub.j,f))−E.sub.C|

d.sub.BS=|E.sub.j,f−E.sub.BS|

[0124] where the addition of σ.sub.t(E.sub.j,f) to d.sub.C is made to take into account energy shift due to the finite resolution of the detector. So peak E.sub.j,f is considered the Compton Edge related to the photo-peak E.sub.i,f and it is discarded if the following relation are true:

d.sub.C<1.0.Math.σ.sub.t(E.sub.j,f)

H.sub.r<0.08∧(R.sub.r>1.1∨Sig/ch.sub.j<0.25)

[0125] where H.sub.r is the ratio of peak heights (peak j-th over peak i-th) and R.sub.r is the ratio σ.sub.j,f/σ.sub.t(E.sub.j,f). On the other hand, peak E.sub.j,f is considered the backscatter peak related to the photo-peak E.sub.i,f if d.sub.BS<1.5.Math.σ.sub.t(E.sub.j,f) and H.sub.r<0.08∧R.sub.r>1.05

[0126] In this case, the backscatter peak E.sub.j,f is not rejected but its index, j, is stored in memory for the subsequent analysis [13.a)].

[0127] 9) Source identification. At this point a set of peak positions, E.sub.i (i.e. all E.sub.i,f detected at the end of 8), i=1, . . . N, with N=number of peaks previously detected, is available. The algorithm compares these values with the known peaks of gamma sources saved in a library, p.sub.j,h, j=1, . . . M, M=number of known sources, h=1, . . . H, H=number of peaks belonging to the j-th known gamma source, s.sub.j.

[0128] For each s.sub.j, p.sub.j,h is identified with E.sub.i if the latter is the nearest of all E.sub.i to p.sub.j,h and if |E.sub.i−p.sub.j,h|<1.5.Math.σ.sub.i, with σ.sub.i=σ.sub.i,f of the i-th peak. If n peaks are identified, n>0, two parameters, Sig and Sig/ch, are computed as the averages over n of Sig and Sig/ch (described in 3).Then another parameter, χ, is introduced to consider the identification quality of s.sub.j. It is computed as:

[00003]$\chi =\frac{L}{n}.Math.{.Math.}_{\underset{i=0}{p_{j,h}\mathrm{identified}}}^n{\left(\frac{E_{\mathrm{id},i}-p_{j,h}}{p_{j,h}}\right)}^2$

[0129] where E.sub.id,i is the E.sub.i related to the identified known peak p.sub.j,h.

[0130] [14.a)] Finally, s.sub.i gamma source is identified if the following conditions are true:

χ<0.002

Sig/ch≥0.17

n/L≥0.59

[0131] Furthermore, an empirical “confidence value”, CL, is associated with the identified source. It is defined as CL=min (CL.sub.1,CL.sub.2), where:

[00004]${\mathrm{CL}}_1=\begin{array}{c}1\mathrm{if}\chi <0.0002\\ 0.90\mathrm{if}0.0002\le \chi <0.0003\\ 0.80\mathrm{if}0.0003\le \chi <0.0004\\ 0.70\mathrm{if}\chi \le 0.0004\end{array}$${\mathrm{CL}}_2=\begin{array}{c}1\mathrm{if}\stackrel{\_}{S\mathrm{.Math.g}/\mathrm{ch}}\ge 0.30\\ 0.90\mathrm{if}0.25\le \stackrel{\_}{S\mathrm{.Math.g}/\mathrm{ch}}<0.30\\ 0.80\mathrm{if}0.20\le \stackrel{\_}{S\mathrm{.Math.g}/\mathrm{ch}}<0.25\\ 0.70\mathrm{if}\stackrel{\_}{S\mathrm{.Math.g}/\mathrm{ch}}<0.20\end{array}$

[0132] In the end [15.a)] the name, S.sub.j, of the identified source s.sub.j is replaced by the string “Backscatter (possible S.sub.j masked)” if (n−n.sub.BS)/L<0.59, assuming n.sub.BS as the number of backscatter peaks detected in 8) and belonging to s.sub.j.

[0133] 10) Minor source identification. Step 9. is repeated for a total of L times, with L=number of libraries of gamma sources, in descending order or priority. For each iteration, every peak E.sub.id,i, related to each identified source s.sub.j, is removed from the set E.sub.i before the next repetition of the source identification routine [16.a)].The library is made as follows (from left to right: source name, source type and gamma energies in keV between square brackets):

TABLE-US-00004 library 1: Am-241 Industrial [59.5409] Ba-133 Industrial [80.8966, 302.8508, 276.3989, 356.0129, 383.8485] Co-56 Industrial [846.75, 511.0, 1238.26, 1771.40] Co-57 Medical [122.06065] Co-60 Industrial, Medical [1173.228, 1332.492] Cs-137 Industrial, Medical [661.657] Eu-152 Industrial, Medical [121.78, 344.27, 778.9045, 963.38, 1112.076, 1407.95] K-40 NORM [1460.822] Mn-54 Industrial [834.848] Na-22 Medical [511.0, 1274.537] Pu-241 SNM [103.68, 208.0, 332.376] Pu-239 SNM [56.0, 375.05, 413.71] Ra-226 NORM [295.21, 351.92, 609.31, 1764.49] Th-232 NORM [238.63, 338.3, 583.0, 911.07, 968.0, 2614.66] U-235 SNM [143.767, 185.72] U-depleted, natural or LEU SNM [1001.026, 766.361] Y-88 Industrial, Medical [898.042, 1836.07] library 2: Ag-108m Industrial [433.93, 614.37, 722.95] Ag-110m Industrial [657.75, 884.67, 937.48, 1384.27] Au-198 Industrial [411.80] Ba-133 Industrial [80.8966, 302.8508, 356.0129] Ba-140 Industrial [537.38] Be-7 NORM [477.61] Bi-207 Industrial [569.70, 1063.66] Co-56 Industrial [846.75, 1238.26, 1771.40] Co-60 Industrial, Medical [1173.228] Cs-134 Industrial [569.32, 604.70, 795.85] Eu-152 Industrial, Medical [121.78, 344.27, 963.38, 1407.95] Ga-67 Medical [93.31, 184.58, 300.23] Hg-203 Industrial [279.20] Ir-192 Industrial, Medical [295.96, 308.46, 316.51, 468.07] La-138 Industrial, Medical [788.74, 1435.80] La-140 Industrial [487.03, 815.83, 1596.49] Lu-177 Medical [112.9498, 208.3662] Mo-99 Medical [739.50, 140.51] Na-22 Medical [511.0] Nd-147 Industrial [91.11, 531.03] Pb-203 Medical [279.20] Ru-103 Industrial [497.08] Ru-106 Industrial, Medical [511.85, 621.84] Sb-124 Industrial [602.72, 722.78, 1691.02] Sb-125 Industrial [427.89, 600.56, 635.90] Sc-46 Industrial [889.28, 1120.55] Sn-113 Industrial, Medical [391.69] Ta-182 Industrial [1121.28, 1189.04, 1221.42] U-235 SNM [185.72] Yb-169 Industrial, Medical [130.52, 177.21, 197.95] Zr-95 Industrial [724.20, 756.73] library 3: 1-131 Medical [364.48] U-depleted, natural or LEU SNM [1001.026] Zn-65 Industrial [1115.55]

[0134] Merging of gamma (type I detector data) and neutron (type II detector data) source identifications

[0135] Before producing the final identification result to the user, the results of both gamma and neutron source identifications, a) and b), are compared [1.c)].The first one is performed in case of gamma and/or neutron alarm [1.a)] while the latter is performed whenever a neutron alarm occurs [1.b)].

[0136] At the end of a) a set of identified sources, S.sup.γ.sub.m, (m=1, . . . k, where k=number of identified sources) is available. As described in point 9) and 10) above, each source in this set is described by type, CL value, Sig parameter and library number, I. On the other hand, b) returns a neutron source, S.sup.n, with its own CL value as result. S.sup.n can be equal to only one of the following sources: “Cf-252” “Pu”, “U”; “AmBe” and “AmLi” (or equal to “Unknown” if the algorithm fails to identify the source).This algorithm can also distinguish shielded sources [7.b)], more specifically the following case:

[0137] Cf-252 shielded with a neutron moderator.

[0138] Cf-252, Pu and AmBe shielded with Pb.

[0139] The merging of gamma and neutron source identification consists in the following operations:

[0140] 1. validation of AmLi source. This gamma and neutron source is identified correctly only if S.sup.n=“AmLi” and at least two sources among “U-235” “Be-7” and “Cs-137” belong to S.sup.γ.sub.m. In this case [2.c)], “U-235”, “Be-7” and “Cs-137” are removed from the set S.sup.γ.sub.m and S.sup.n is still “AmLi”, otherwise S.sup.n is set to “Unknown” [3.c)].

[0141] 2. masking condition determination for neutron sources and warning about S.sup.n masked with gamma source. The word “masked” is appended to S.sup.n name if S.sup.n is not “Unknown” and there is at least one gamma source in S.sup.γ.sub.k that is different from S.sup.n [4.c)].

[0142] 3. identification of Na-22/Co-60 with a higher counting rate of Co-60/Na-22. It is possible that only 1 out of 2 peaks of Na-22/Co-60 is detected when one of these gamma sources is more intense than the other one, because of the finite resolution of the detector. For this reason, Na-22 and Co-60 have been added also in the 2.sup.nd library with only one peak (511 keV and 1173 keV for Na-22 and Co-60, respectively). In this way both the two sources will be identified when Na-22/Co-60 is identified from the 2.sup.nd library and Co-60/Na-22 from the 1.sup.st one [7.c), 8.c)]. Possible Ru-103 source in S.sup.γ.sub.m is removed from the identified sources when Na-22 and Co-60 are identified from the 2.sup.nd and the 1.sup.st library, respectively.

[0143] 4. estimation of Pu-239 enrichment grade. This step is performed if at least one isotope of Pu, Pu-n, is identified by a) and no other gamma source, with gamma peaks between 180 and 243 keV, belongs to S.sup.γ.sub.m (according to the current library: Ga-67, Lu-177, Th-232, U-235, Yb-169). In this case [5.c)], Pu-239 enrichment grade is defined by the following 5 classes, based on the ratio P.sub.208/B.sub.208 that has been computed by the gamma source identification algorithm (point 5) above):

[0144] i. Weapon Grade Pu (WGPu) if S.sup.n=Pu, “Pu-239” belongs to S.sup.γ.sub.m and P.sub.208/B.sub.208<0.11. The word “WGPu” is appended to the source name of S.sup.n.

[0145] ii. 80-90% enriched if “Pu-239” belongs to S.sup.γ.sub.m and 0.11≤P.sub.208/B.sub.208<0.18. The string “80-90% Pu-239” is appended to the source name of S.sup.n.

[0146] iii. 65-80% enriched if 0.18≤P.sub.208/B.sub.208<0.31. The string “65-80% Pu-239” is appended to the source name of S.sup.n.

[0147] iv. 55-70% enriched if 0.31≤P.sub.208/B.sub.208<0.45.s. The string “55-70% Pu-239” is appended to the source name of S.sup.n.

[0148] v. <60% enriched if P.sub.208/B.sub.208≥0.45. The string “<60% Pu-239” is appended to the source name of S.sup.n.

[0149] Final enrichment grade is reduced by 10% in two cases [6.c)]: if S.sup.n is identified as a shielded source or if the ratio Sig.sub.γ/Sig.sub.Pu is smaller than 1.0, assuming Sig.sub.γ as the maximum value of Sig parameter among all gamma sources in S.sup.γ.sub.m (excluding Am-241 and Pu-n) and Sig.sub.Pu as the corresponding value for Pu sources in S.sup.γ.sub.m.

[0150] The present invention also includes the identification of SNM in conditions of scenarios with the presence of shielding and masking, i.e., typically scenarios that could be considered in potential terrorist action or in the presence of nuclear “smuggling” situations.

[0151] The system analyzes in real time like in a chart, where the ratio R.sub.c between the gamma and neutron counts is shown in the x-axis, and the ratio R.sub.E between the average energy of the gammas and that of the neutrons is shown in the y-axis.

[0152] In this way it is possible to precisely determine the positioning of the source under observation in a rather well confinable space and therefore associable with that specific situation, as in the case of the enclosed FIG. 5, where the grouping or cluster of the points relating to Cf252 measurements can be seen without any presence of shielding and masking in well-defined areas which have been obtained. FIG. 6 shows the positioning of the points in the presence of lead shielding, FIG. 7 shows the presence of polyethylene (PE) neutron moderator shielding and FIG. 8 shows the presence of different sources to create masking scenarios for the neutron source which in this case are resolved by the method and system according to the present invention by virtue of the simultaneous spectroscopic evaluation. Enclosed FIG. 9 to FIG. 17 summarize the experimental observations using the algorithm for SNM identification according to the present invention.

[0153] FIG. 9 shows the plane distribution of all the identifications (data: ENEA, Legnaro, Seibersdorf and measurements with AmLi, Oak Ridge). In each chart, the ratio R.sub.c between the gamma and neutron counts is shown in the x-axis, and the ratio R.sub.E between the average energy of the gammas and the neutrons is shown in the y-axis. These values are obtained during the identification, with the data obtained by the liquid scintillator (EJ-309) in the identification measurement (in this specific case of 1 min). In particular, the last saved spectrum (in this specific case of 3 min acquisition) of the ambient background, appropriately rescaled, is subtracted from the gamma energy spectrum.

[0154] The boxes shown in the charts identify the regions in the plane R.sub.C, R.sub.E in which to identify the different source types. For the identification of AmBe, since the relative box overlaps that of Cf-252 and Pu, the average neutron energy is used as the third parameter (AmBe if E.sub.n>650 keV.sub.ce).

[0155] In the chart in FIG. 10, in fact, the measurements with a number of neutrons >8 cpm were selected. In this case, although the set threshold is less restrictive than that due to the alarm on the neutron cps (generally the alarm sounds when the neutrons exceed ˜0.2 cps=12 cpm), the various sources are well separated in the corresponding regions. We can therefore say that the algorithm is also effective for low neutron counts, compatible with the background, and that, at present, the neutron alarm constitutes a more than sufficient condition for the correct recognition of a source. For example, in the case of U sources, it can be seen how the algorithm correctly identified the source including that made in Seibersdorf with HEU (90% U-235).

[0156] We can also observe the position of the 2 measurements with AmLi, placed between the boxes of Cf-252 and U.

[0157] To better highlight the behavior of the algorithm with the different sources, FIG. 11 shows the results obtained with the sources only, without shielding material, with a number of neutrons >8 cpm.

[0158] From this chart the regions of the plane occupied by the different sources are evident.sup.2. Based on these data, regions have been defined with confidence level (CL)=100%.

[0159] In the following charts, using the same conditions, FIG. 12 shows the Pu sources divided between CBNM side A and side B (ENEA—side A and B have different gamma emissions because the housing of the source has a side with thinner shielding) and FIG. 13 shows the Pu sources with different enrichment levels of Pu-239 (Seibersdorf).

[0160] From what can be seen in the chart in FIG. 13, also considering the absence of shielding and external masking sources, the well-defined positioning of the Pu with different levels of enrichment can be seen.

[0161] FIG. 14, FIG. 15 and FIG. 16 show the various measurements with shielding, respectively only Pb (5 cm), only PE (from 2 cm to 10 cm) and Pb+PE (Pb 5 cm, PE from 2 to 10 cm), always keeping a number of neutrons >8 cpm (0.133 cps).

[0162] From these charts the expected behavior is verified: the presence of Pb shifts the measurement towards the regions with less R.sub.C, unlike polyethylene.

[0163] Finally, FIG. 17 shows cases with gamma sources in addition to the neutron source (Cf-252 in blue, Pu in orange). In the case of the Cf-252, it can be seen how the points move in the plane according to the energy of the gamma sources used as masking (in addition to their relative dose). As the dose due to the gamma masking source increases, the points move instead to the regions with greater R.sub.C (increasing range multiplicity). In this way the regions were defined to identify the presence of masking.

[0164] With reference to enclosed FIG. 18 a portable device for the recognition of a radioactive source is generally indicated with 1. A radioactive source Rad, which emits both n neutrons and gamma γ radiation, is schematically indicated.

[0165] The portable device 1 is equipped with a type II detector 3 consisting, in a preferred embodiment, of a cylindrical cell of organic liquid scintillator. In another embodiment the detector 3 is of the type that uses xylene as an organic solvent, i.e., the EJ 309 model by Eljen Technology which has innovative characteristics like low flammability and lower toxicity which make it suitable for commercial use on a commercial scale.

[0166] The sensor is coupled to a photomultiplier or SiPm readout 4, for example the Hamamatsu model H12700, connected to a miniaturized power supply circuit 6 that provides to the photomultiplier 4 the high voltage necessary for its correct operation.

[0167] Detector 3 can detect both the light pulses produced by the interaction with neutrons (interaction known as “n-p scattering”) and the gamma radiation (interaction known as “Compton scattering”).

[0168] A digitizer circuit 5 of the fast type is adapted to convert light pulses into digital signals, analyzed and processed in line through a programmable integrated digital processor 9, e.g., a FPGA (“Field Programmable Gate Array”).

[0169] The portable device 1 is also equipped with a type I detector 3-bis consisting of a cylindrical cell of organic crystal scintillator. In a preferred embodiment variant, the sensor 3-bis is of the type that uses Cerium-Bromide (CeBr.sub.3) crystal lattice, i.e., the CeBr.sub.3 model by Scionix-Holland which provides a greater ability to discriminate gamma γ radiation close in energy (this characteristic is called detector resolution: a better resolution provides higher ability to distinguish two gamma γ close in energy).

[0170] The sensor 3-bis is coupled to a photomultiplier tube or SiPm readout 4-bis, connected to a miniaturized power supply circuit 6-bis that provides to the photomultiplier 4-bis the high voltage necessary for its correct operation.

[0171] Sensor 3-bis can detect only the light pulses produced by the interaction with gamma radiation γ while its sensitivity to neutron n radiation is close to zero. The sensor 3-bis interact with gamma not only with interaction known as “Compton scattering” (like sensor 3) but also with interaction known as “photoelectric effect”. This effect releases all the energy of the gamma γ radiation in a single interaction and the detector can record the total energy of the gamma in a single shot. Since each Radioactive source Rad has its own characteristic gamma γ lines (each one with different energies), by collecting all the energy of each gamma is possible to exactly recognize the radioactive source Rad (like with a fingerprint) with a single measurement made with the sensor 3-bis.

[0172] The light pulses (fingerprint) collected by the sensor 3-bis are converted, by means of a digitizer circuit 5-bis of the fast type, into digital signals, analyzed and processed in line through a programmable integrated digital circuit 9, e.g., a FPGA (“Field Programmable Gate Array”).

[0173] The device 1 also embeds a graphic user interface (GUI) 7 and a control panel 8 for the exchange of information and data between the device 1 and its user.

[0174] All components of the device 1 have dimensions and weights compatible with a single user transportation and manipulation.

[0175] The programmable integrated circuit 9 may works also as an internal memory of the device 1, in which a data processing program comprising a plurality of instructions for the execution of the method according to the present invention is loaded. Said method allowing the recognition of the radioactive source Rad when the program is executed from portable device 1.